biological basis of radiotherapy: where do we stand? · 2014-09-11 · 1 biological basis of...
TRANSCRIPT
1
Biological basis of radiotherapy:
where do we stand?
International workshop
Stockholm, 4-5 September 2014
Program and book of abstracts
The Swedish Royal Academy of Sciences,
Lilla Frescativägen 4A
114 18 Stockholm, Sweden
(www.kva.se)
3
About the workshop
With a history of more than 100 years, radiation therapy remains one of the main modalities
used in the management of cancer together with surgery and chemotherapy.
The progress in treatment planning, image guidance and radiation delivery has led to the
appearance of high precision radiotherapy that is a common feature in many clinics.
Furthermore, the technological development of functional and molecular techniques for
imaging the tumours has opened new possibilities for defining the target and devising the
treatment in an innovative manner.
However important questions remain with respect to the relevant clinical and radiobiological
aspects.
For radiobiology in particular, progress in research is not accompanied by a quick clinical
implementation in spite of its translational character.
Classical radiobiology with its famous 5 R’s and the linear-quadratic model for clonogenic
survival has been the most influential component of the radiotherapy fractionation schedule
design and calculations of isoeffects, while some modern findings do not easily find their way
from bench to bedside.
This workshop aims to revisit the old school of radiobiology and identify new findings that
have potential to impact on the clinical practice and lead towards the next big leap in clinical
radiotherapy: the development of high precision individualised radiotherapy.
Organisers: Iuliana Toma-Dasu Medical Radiation Physics, Stockholm University and Karolinska
Institutet
Andrzej Wojcik Centre for Radiation Protection Research, Stockholm University
Emely Lindblom Scientific secretary - Medical Radiation Physics, Stockholm University
and Karolinska Institutet
Venue: The Swedish Royal Academy of Sciences, Lilla Frescativägen 4A
114 18 Stockholm, Sweden (www.kva.se)
4
Invited speakers:
Jan Bussink - Radboud University Nijmegen Medical Centre, Nijmegen
Roger Dale - Department of Surgery and Cancer, Faculty of Medicine, Imperial College, London
Alexandru Dasu- Department of Radiation Physics UHL, Linköping University, Linköping
Anna Dubrovska - OncoRay Center for Radiation Research in Oncology, Dresden, Germany
Marco Durante - GSI Helmholtzzentrum für Schwerionenforschung, Darmstad
Eva Forssell-Aronsson - Sahlgrenska University Hospital, Gothenburg
Jack Fowler - University of Wisconsin Medical School, Madison, Wisconsin
Ester Hammond - Cancer Research UK and Medical Research Council Oxford Institute for
Radiation Oncology, Oxford
Jolyon Hendry - Christie Hospital, Manchester
Carsten Herskind – Department of Radiation Oncology, Universitaetsmedizin Mannheim,
Medical Faculty Mannheim, Heidelberg University, Mannheim
Michael Joiner – Department of Radiation Oncology at Wayne State University School of
Medicine in Detroit, Michigan
Wolfgang Sauerwein – University Clinics Essen, Essen
Albert Siegbahn – Medical Radiation Physics, Stockholm University and Karolinska Institute
Klaus Trott - University of Pavia, Italy
Conchita Vens - Experimental Therapy Division at the Netherlands Cancer Institute, Amsterdam
5
PROGRAM
September 4, 2014
Introductory lecture
08:30-09:00 Radiobiology of clinical fractionated
radiotherapy – Textbook versus new
knowledge
Iuliana Toma-Dasu
Session 1 - What is the target of radiotherapy?
Chairperson: Rolf Lewensohn
09:00-09:40 The importance of tumour stem cells for
radiotherapy
Anna Dubrovska
09:40-10:20 The importance of normal tissue stem cells
for radiotherapy
Klaus Trott
Coffee break
10:40-11:20 The importance of tumour environment for
radiotherapy
Jan Bussink
11:20-12:00 The development of beam-grid based
radiosurgery
Albert Siegbahn
Lunch break
Session 2 - The classic 4 Rs of radiotherapy – are they still valid?
Chairperson: Mats Harms-Ringdahl
13:00-13:40 Repair of sublethal damage in tumour and
normal tissue cells
Conchita Vens
13:40-14:20 Tumour cells reassortment within the cell
cycle (including check points and cell cycle
arrest)
Carsten Herskind
Coffee break
14:40-15:20 Proliferation and accelerated repopulation in
tumour and normal tissue
Jolyon Hendry
15:20-16:00 Hypoxia, reoxygenation and radiation
sensitivity
Ester Hammond
Coffee break
16:30-17:30 General discussion Moderator: Claes Mercke
19:00 Dinner at Karolinska Hospital, CCK building, floor 5 (see maps). Busses will
be provided to transport participants to Karolinska.
6
September 5, 2014
Session 3 - The LQ model and its parameters
Chairperson: Per Nilsson
09:00-09:40 Radiobiological basis of the LQ model Mike Joiner
09:40-10:20 The radiobiological modelling challenges of
21st century radiotherapy
Roger Dale
Coffee break
10:40-11:20 LQ parameters – Does one size fit all?
Heterogeneity in parameters versus one
single set of parameters for all the cells in the
tumour and normal tissue
Alexandru Dasu
11:20-12:00 Is there an optimal treatment time and
fractionated schedule?
Jack Fowler
Lunch break
Session 4 – New/old treatment modalities
Chairperson: Bo Stenerlöw
13:00-13:40 Biological basis of targeted radiotherapy Eva Forssell-Aronsson
13:40-14:20 Biological basis of brachytherapy Andrzej Wojcik
Coffee break
14:40-15:20 Biological basis of hadron therapy Marco Durante
15:20-16:00 Biological basis of neutron therapy Wolfgang Sauerwein
Coffee break
16:30-17:30 General discussion and concluding
remarks
Moderator: Mike Joiner
17:30 – end of meeting
7
Invited speakers
The importance of the tumor microenvironment for radiotherapy
Jan Bussink
Department of Radiation Oncology (874)
Radboud university medical center
P.O. Box 9101, 6500 HM Nijmegen
E-mail: [email protected]
The microenvironment consists of a complex symbiosis of various closely interacting cells.
For growth, the tumor is depending on nutrients and oxygen from the ‘host’ system. Also,
waste products are discarded towards that same ‘host’. Sensitivity and resistance to
radiotherapy is independently connected to the microenvironment. The complex and often
chaotic vascular architecture, induced by tumor cells itself, leads to heterogeneous
distribution of hypoxic cells, proliferating cells immune cells etc. Consequently insufficient
oxygenation of tumor cells makes them more resistant to ionizing radiation. Treatment by
irradiation not only depends on the microenvironment, it also severely affects the
microenvironment. Radiotherapy leads to changes in the balance with respect to vascular
damage and vascular remodeling, hypoxia and reoxygenation, immune suppression and
immune activation. The position of this balance contributes to the faith of a tumor.
8
The radiobiological modelling challenges of 21st century radiotherapy
Roger Dale
Department of Surgery and Cancer
Faculty of Medicine
Imperial College
London, UK
Radiotherapy remains a successful and cost-effective treatment modality for a wide range of
cancers and, despite occasional suggestions of its imminent demise, new and exciting
techniques continue to be developed. Many of the newer treatments represent considerable
departures from the types of radiotherapy which have been practiced for over 100 years,
meaning that the bedrock of previous clinical experience may not always be relevant in
guiding the optimal application of the new methods. The requirement for a sound
radiobiology base from which to quantitatively guide the optimal application of the emergent
technologies is thus stronger than ever before.
This talk will review some of the challenges posed by new radiotherapy practices and will
highlight those areas where current radiobiological understanding is inadequate or in need of
further refinement. Even in cases where current radiobiological models are known to be
inadequate descriptors of events at the macroscopic (i.e. clinically observable) level, it is
important not to lose sight of the fact that, at the microscopic level, the bio-physical origins of
the models may be sound. This suggests that translational research has a major role to play in
adapting radiobiology to the requirements of 21st Century treatment practice and that there is
an urgent need to promote and support such research and disseminate and implement
findings.
To cultivate an appreciation of why such research is required there is a parallel requirement
for the next generation of radiation oncologists and radiation scientists to understand both the
limitations and potential of radiobiology and, for this reason, on-going education programmes
in radiobiology will be of key importance.
9
LQ parameters for modelling - Does one size fit all?
Alexandru Dasu
Department of Radiation Physics and Department of Medical and Health Sciences, Linköping
University, Linköping, Sweden
LQ modelling with parameters derived from populations of patients has been used for many
years for accurate isoeffect calculations regarding the response of tumours or the tolerance of
late and acute reacting normal tissues from many clinically-used fractionated schedules. The
parameters used are thought to describe the “average patient”, but in reality rather broad
interval of values may be found in individuals in the investigated populations and therefore
the value of pre-treatment derived parameters to predict treatment response has often been
debated. Furthermore, clinical outcome can often be modelled by assuming quite extreme
single values for the parameters, while experimentally-derived parameters require
assumptions about inter-patient heterogeneity as heterogeneity in radiation sensitivity or
tumour cell population influences the position and the slope of the tumour control probability
curve and therefore directly impacts upon predictions for treatment outcome. This lecture
aims to explore the question of parameters to be used for LQ modelling, the opportunity of
using parameters relevant to individuals or the average, as well as the impact of accounting
for inter-patient heterogeneity in tumour control probability predictions or the search for an
optimum overall treatment time.
10
Radiotherapy and cancer stem cells
Anna Dubrovska
OncoRay – National Center for Radiation Research in Oncology, Medical Faculty and University
Hospital
Carl Gustav Carus Technische Universität Dresden, Helmholtz-Zentrum Dresden-Rossendorf
Dresden, German Cancer Consortium (DKTK) Dresden, and German Cancer Research Center
(DKFZ) Heidelberg, Germany
Radiotherapy is a curative treatment option in many types of tumors. Nevertheless, cancer
patients can relapse after radiation therapy. The radiotherapy failure might be related to
cancer stem cell (CSC) population which was not completely sterilized during the treatment
and can provoke tumor re-growth [1-3]. The proportion of CSCs has a high intertumoral
variability. Current findings suggest that estimation of the number of CSCs in pre-therapeutic
tumor biopsies might be predictive of tumor radiocurability in some types of cancer including
cervical squamous cell carcinoma (CSCC), head and neck squamous cell carcinoma
(HNSCC), rectal carcinoma and glioma patients [2, 3]. Despite some limitations of these
studies, they have paved an avenue for future implication of CSC-related predictive
biomarkers for radiotherapy. In addition to the impact of CSC density on tumor
radiocurability, recent experimental reports suggest a number of different intrinsic and
extrinsic adaptations that confer tumor radioresistance and which also occur in CSC
populations, including quiescence, increased DNA repair capability, activation of the
cell survival pathways and scavenging of the reactive oxygen species (ROS) that can
induce DNA damage [3-5]. A broad variety of the microenvironmental niches as well as
genetic and epigenetic changes of tumor cells during tumor development make cellular
radiosensitivity dynamic in nature [3, 5-7]. The talk will review the role of CSCs in tumor
radiocurability and their implication in the development of predictive assay for radiotherapy.
The limitations of the current models to study CSC properties including their putative
radioresistance will be also discussed.
References
1. Baumann M, Krause M, Hill R. 2008. Exploring the role of cancer stem cells in
radioresistance. Nat Rev Cancer 8: 545 – 554.
11
2. Bütof R, Dubrovska A, Baumann M. 2013. Clinical perspectives of cancer stem cell research
in radiation oncology . Radiother Oncol. 108: 388 – 396.
3. Peitzsch C, Kurth I, Kunz-Schughart L , Baumann M, Dubrovska A. 2013.
Discovery of the cancer stem cell related determinants of radioresistance. Radiother Oncol
108: 378 – 387.
4. Pajonk F, Vlashi E, McBride WH. 2010. Radiation resistance of cancer stem cells: The 4
R’s of radiobiology revisited. Stem Cells 28: 639 – 648.
5. Rycaj K, Tang DG. 2014. Cancer stem cells and radioresistance. Int J Radiat Biol 90:
615– 621.
6. Tang DG. 2012. Understanding cancer stem cell heterogeneity and plasticity. Cell Res 22:
457–472.
7. Kreso A, Dick JE. 2014. Evolution of the Cancer Stem Cell Model. Cell Stem Cell 14: 275 –
291.
12
Biological basis of hadron therapy
Marco Durante
GSI Helmholtzzentrum für Schwerionenforschung, Biophysics Department, Darmstadt, Germany
The use of high-energy charged particles in radiotherapy was proposed several years ago,
both for their physical and radiobiological advantages compared to X-rays. Particle therapy is
now generally considered a cutting-edge technology with a tremendous potential for several
cancers associated to poor prognosis, such as pancreas, locally recurrent rectal cancer, and
lung. Protons and carbon ions have been used for treating many different solid cancers, and
several new centers with large accelerators are under construction. Meanwhile, tremendous
technological improvements in image-guided radiation therapy (IGRT) paved the way to
hypofractionation (down to high-dose, single fraction treatments) of X-rays for both cranial
(stereotactic radiosurgery) and extracranial (stereotactive ablative radiotherapy) targets.
These new technologies can be competitive with surgery, for example in lung cancer. Can
particle therapy produce better results in radiosurgery and eventually become an alternative to
surgery? The talk will try to address this question and present research challenges in medical
physics and radiobiology.
13
Optimum Overall Times for Advanced Head and Neck Radiotherapy
Jack F. Fowler
Emeritus Professor, Retired from the Department of Human Oncology, University of Wisconsin
Medical School, Madison, WI, U.S.A.
Radiotherapy has a history of more than 100 years of the impressive record of being second
only to surgery in the treatment of cancers. It is a non-invasive technique that relies on
improving the localisation of the target, the dose delivery to smaller and smaller volumes and
on detecting differences in biological properties, including genetic properties, to increase the
outcome. Nevertheless, a central issue of modern radiotherapy remains finding the best
optimal schedules that can be obtained from radiation itself. Two different constraints govern
any radiation oncology treatment, one that tries to make the overall time longer and the other
that tries to make it shorter. The natural constraint that tends to lengthen the overall treatment
time is the successive addition of a small dose per day of treatment, which might be 2 Gy per
fraction per treatment day for example. The constraint for limiting the length of the overall
time comes from the repopulation rate of the malignant cells of the tumour during the
treatment. It is from balancing these two opposing forces acting on the treatment that leads to
an optimum overall time for the radiotherapy of rapidly proliferating tumours like those of
the head and neck, keeping also in mind the constraints set by the normal tissues. Thus, the
dose to the late normal tissues must not exceed the total tolerance dose equivalent to 70 Gy in
35 fractions, while acute mucosal reactions have a total tolerance dose equivalent to 51.3 Gy
in 2 Gy fractions. Several strategies have been proposed to fulfil these constraints including
delivering one or several fractions per day. This lecture focuses on obtaining and exploring
the best optimal schedules in modern radiation therapy of head and neck cancers. Although it
has become clear human papilloma virus cancers do better when treated with radiotherapy
than those not infected with the virus that does not alter the necessity of using the best
radiotherapy schedule that can be obtained with respect to its optimum overall time. This
makes the search for the optimum overall times an actual problem of modern radiotherapy.
14
Targeting hypoxic tumour cells through the DNA damage response
Ester M. Hammond
The CR-UK/MRC Oxford Institute for Radiation Oncology, University of Oxford, Oxford, OX3
7DQ, UK
Tumour hypoxia is a common feature of solid tumours and negatively impacts patient
prognosis. The increased radioresistance of hypoxic cells significantly decreases the efficacy
of radiotherapy. The most severely hypoxic regions of a tumour are also the most
radioresistant (radiobiologic hypoxia). In response to radiobiologic hypoxia a DNA damage
response (DDR), including ATM and ATR, is induced. Our studies have demonstrated that
inhibition of key factors in the DDR, for example ATR, ATM and Chk1, can sensitise cells to
hypoxia and subsequent reoxygenation. Most importantly, we have shown that a number of
DDR inhibitors increase the radiosensitivity of hypoxic cells. These combinations offer new
strategies to target tumour hypoxia. Most recently, we have developed a series of hypoxia-
activated prodrugs. These agents are inactive at normal oxygen levels but in conditions of
low oxygen are reduced to release an active inhibitor. The targeting of Chk1 inhibition to
hypoxic cells will be discussed.
15
Repopulation concepts in irradiated normal tissues and tumours
Jolyon Hendry
Christie Medical Physics and Engineering, Christie Hospital NHS Foundation Trust, Manchester, UK.
Stem cells are the best cells you will ever have to repopulate depleted normal tissues.
Lineages of stem cells, daughter differentiating cells (transit cells) and mature functional cells
have been characterised in many tissues. Some lineages are short (breast, skin), and others
long (bone marrow, spermatogenesis). The Labelling Index in these tissues is governed
mainly by the predominant dividing transit cells, not by the stem cells. Up to 6 surface
biomarkers have been found necessary to identify stem cells in various tissues. The stem-cell
self-renewal probability (p) is 0.5 in steady-state (no net growth), and about 0.6 in
exponential growth after much radiation-induced depletion. During 2 Gy daily fractionation,
there is a latency interval followed by repopulation equivalent to about 1 Gy/day (Dprolif) in
human oral mucosa.
In untreated human tumours, a short volume doubling time of 40 days could be explained by
a cell cycle time of 2 days, a growth fraction of 50% (giving a potential doubling time of 4
days), and a cell loss factor of 90%, if all tumour cells were malignant. During 2 Gy daily
fractionation, Dprolif is 0.6-0.8 Gy/day for head & neck tumours. This has led to guidelines for
avoiding detrimental interruptions in treatment, and to new trials of accelerated schedules.
There has been renewed interest in the concept of cancer stem cells (CSC), where tumours
retain some of the lineage characteristics of their tissue of origin. This comes from the
histological grading of differentiation status, and from many experimental studies of target
cell number and other characteristics. The concept suggests that well-differentiated tumours
may show some of the accelerated repopulation characteristic of normal tissues and should
benefit more from accelerated schedules, whereas longer conventional schedules could be
used for poorly-differentiated tumours. Also, it opens up new developments based on target
cell imaging linked to dose distributions, and combined treatments to overcome CSC
resistance. Identification of more CSC biomarkers would help these approaches.
16
Tumor cell reassortment within the cell cycle (including checkpoints and cell
cycle arrest)
Carsten Herskind
Department of Radiation Oncology, Universitaetsmedizin Mannheim, Medical Faculty Mannheim,
Heidelberg University, Mannheim, Germany
e-mail: [email protected]
The survival advantage of cells irradiated in radioresistant cell-cycle phases makes
reassortment of the surviving cells over the more radiosensitive cell-cycle phases part of the
rationale for fractionated radiotherapy. While it has not been possible to exploit the partial
synchronization of tumour cells in clinical practice, understanding the regulation of cell-cycle
progression after irradiation may nevertheless be important for changes in fraction size and
may potentially be used to manipulate the radiosensitivity of cancer cells. The DNA damage
response involves cell-cycle arrest at different checkpoints, repair of damage, as well as the
decision to resume proliferation or prevent cell reproduction if residual damage cannot be
repaired or is misrepaired. Thus checkpoints are present before and during replication in the
G1/S and S phases, and before mitosis in the G2/M phase. However, novel sophisticated
techniques in cell-cycle labeling and signal transduction have revealed several molecularly
distinct checkpoints differing in kinetics, efficiency, and dose dependence, which can be
suppressed by small molecule inhibitors or by RNA interference. Thus, the most conserved
checkpoint in G2/M has been shown to be regulated by at least three different pathways. The
current understanding of the DNA damage checkpoints in S, G2/M, and G1/S, will be
reviewed with special emphasis on their sensitivity, imperfections and the release of cells
from the various checkpoints. Furthermore the potential for modulating radiosensitivity will
be discussed.
17
Radiobiological basis of the LQ model – Is there one…?
Michael C. Joiner
Department of Radiation Oncology, Wayne State University School of Medicine, Detroit, MI
Biological effect and physical dose are non-linearly related. This was implicitly recognized
very early in the development of radiotherapy, and led to the power-law equations which
dominated thinking for several decades. Deficiencies in the power-law description were
identified even as early as the 1960’s, and in the 1970’s and 80’s a massive amount of
experimental work on many newly-developed models of normal-tissue injury in rodents and
pigs, showed that –ln(Surviving Fraction) is much better described by a second-order
polynomial in dose, the well-known Linear-Quadratic (LQ) equation. The LQ description has
since been thoroughly tested in the clinical domain. LQ is the simplest mathematical
description of a non-linear relationship and though empirical in nature, it has nevertheless
been subject to many attempts to connect with our understanding of how radiation injury is
produced and repaired at the cell and molecular level. Yet any meaningful and clinically
useful link in this respect has remained largely elusive.
LQ deals specifically with the relationship between total dose and dose per fraction, and
with interfraction interval using the Incomplete-Repair derivative model. The relationship
between total dose and overall treatment time, originally included in the power-law models in
a manner that didn’t match actual clinical observations, is an even more complex relationship
dependent on the different underlying radiobiology of different tissues even within the
apparently same category of early-reacting or late-reacting tissues, distinguished by
respectively a “high” or “low” ratio of α/β in the LQ equation. Overall time is therefore better
handled independently of LQ. Prediction of tolerance to retreatment is an even more complex
issue where our radiobiological understanding is earlier in development and useful models
are not yet to hand.
A straightforward but untested hypothesis for the different α/β values for early- and late-
reacting tissues, is that a naturally low α/β for a target cell population is smoothed out to a
higher value as the sum of the responses of different proliferative subpopulations, and
different phases of the cell cycle that these are in. This same explanation could also be
applied
18
to the responses of malignancies in the lung and head and neck, also adding in the additional
response variation of cells at various levels of hypoxia in these sites. Of note is the recent
connection made between outcome of radiotherapy and HPV status in oropharyngeal cancers,
which implies a possible difference in treatment strategy between these tumor subtypes and
could also explain the high α/β of head and neck cancer overall as the sum of the responses of
the different cancer subtypes (HPV + and −) which could both have low α/β but different
radiosensitivity. Importantly, it is now evident that in some malignancies, notably prostate
and breast, clinical data do indeed indicate a low α/β which might also reflect more
uniformity in response perhaps more characteristic of lower proliferative or early-stage
disease. This has resulted in new efforts to test hypofractionation which have also been
enabled by the better dose localization achievable with image-guided Volumetric Modulated
Arc Therapy.
There is evidence that the LQ model becomes less reliable at doses per fraction < 1 Gy, due
to possible low-dose hyper-radiosensitivity, and also at > 6 Gy per fraction for reasons not
yet understood. Though, it is axiomatic that LQ must indeed overestimate effect at very high
doses per fraction because the effective D0 would become unrealistically low. This makes the
outcome of hypofractionated regimes less predictable: using LQ at high doses per fraction
would be playing safe in predicting toxicity of hypofractionation, while overestimating the
effect on the target malignancy, noting that possible hypoxia in a tumor could also limit the
effectiveness of large dose fractions.
19
The development of a beam-grid based radiosurgery
Albert Siegbahn
Medical Radiation Physics, Stockholm University and Karolinska Institutet
At the onset of the lecture, the historical development of radiation grid therapy will be
presented. Thereafter, the dose volume effect and the potential benefit of using beam arrays
with small beam elements will be discussed. The in vivo microbeam experiments with rodents,
undertaken for the first time in the 1950s as a part of the United States space exploration
program, showed that there is nearly no observable functional damage of central nervous
system tissue after such irradiations. In an attempt to explore this large tissue tolerance to
microscopic beams, x-ray microbeam-grid radiation therapy preclinical trials began in the
synchrotron laboratory at Brookhaven National Laboratory in the late 1990s. The idea was to
irradiate a cancer target from different directions up to high doses with spatially-fractionated
x-ray beams. At this point in time, nearly parallel x-ray beams of sufficiently high energy (for
depth penetration) and fluence (to avoid motion issues) had become available which was
deemed necessary for this kind of treatment. Narrow, micrometer-wide x-ray beams with a
close separation form the beam array used for this therapy. In recent years, this method has
been developed and tested at different synchrotron laboratories on four different continents.
To date, MRT has been evaluated mainly with small experimental animals, e.g. rodents, some
of them bearing a transplanted tumour. Data from these tests will be presented in this lecture.
MRT preclinical trials ultimately aiming toward treating humans have recently moved forward
with treatments of larger animals, e.g. cats, with spontaneous tumours. Some tests have also
been performed by different groups with beam arrays containing other types of particles, such
as with grids of proton or carbon-ion beams. With ions the additional possibility exists to spare
sensitive structures posterior to the target in the beam direction.
20
The importance of normal tissue stem cells for radiotherapy
Klaus-Rüdiger Trott
Università degli Studi di Pavia
The existence and the radiosensitivity of normal tissue stem cells and their role in the
pathogenesis of normal tissue damage was first demonstrated in the bone marrow. Fatal bone
marrow hypoplasia and subsequent agranulocytosis is related to sterilization of about 99.9%
of bone marrow stem cells. The clinical consequences can be avoided by intravenous
injection of a sufficient number of viable bone marrow stem cells into the irradiated animal,
which not only gives additional proof of the role of bone marrow stem cell sterilization in the
pathogenesis of bone marrow failure but also was the basis of clinical application of stem cell
transplantation in the treatment of radiation-induced hypoplasia. Extrapolating these
radiobiological data to all other normal organs and tissues, inactivation of normal tissue stem
cells (called tissue rescuing units) was widely taken as the main pathogenic pathway leading
to normal tissue damage from radiotherapy. However, experimental and clinical radiobiology
research demonstrated that whereas in early normal tissue damage of some tissues such as
oral mucositis and intestinal mucositis, radiation damage of stem cells play some role, the
pathogenesis of late normal tissue damage is much more complex and stem cells appear to
play no significant role. Each type of late complication after radiotherapy depends on its own
specific mechanism which is triggered by the radiation exposure of particular structures or
subvolumes of the respective organ at risk. The radiobiological mechanisms which are
involved in the resulting pathogenesis differ between the different complications, even in the
same organ which follow different dose effect relationships and show different dependence
on dose per fraction, overall treatment time and anatomical dose distribution within the
subvolume at risk. The different mechanisms can be classified as:
1. single cell effects such as “cell death”, sterilisation of stem cells, inhibition of
proliferation,
2. tissue effects which depend on the interaction between different cells and cell
populations within organs or between organs such as inflammation or differentiation,
3. effects which result from alterations of tissue structure and function such as vascular
injury,
21
4. other, less well defined functional changes such as alterations in neuromuscular
function or immunological responses.
The most important role normal tissue stem cells are likely to play in the pathogenesis of late
complications of radiotherapy is related to the induction of second cancers, however even
there their role is controversially discussed.
This work was performed as part of the FP7 supported research project ALLEGRO.
22
Biological basis of brachytherapy
Andrzej Wojcik
Stockholm University
Brachytherapy is the delivery of radiation therapy using sealed sources that are placed as close
as possible to the site to be treated. The major advantages of brachytherapy lie in the physics
of dose distribution around the source, which results in a high concentration of dose around the
source. A disadvantage lies in the need to access the tumour with an operative procedure
requiring very skilled personnel to undertake the treatment.
Brachytherapy is most often used as boost to external beam radiotherapy in order to increase
the dose to the tumour without increasing the dose to normal tissues. This approach is
understandable and it is often claimed that the radiobiology of brachytherapy is governed by
exactly the same principles as external beam therapy (the 4 Rs of radiotherapy).
In the treatment of prostate cancer brachytherapy is increasingly used as monotherapy. This
became possible thanks to the development of trans-rectal ultrasound (TRUS) to precisely
position the radiation sources. Initially, the monotherapy was used to treat low and intermediate
risk group patients with I-125 seeds yielding a low dose rate (LDR). At the same time it became
increasingly evident that the α/β value for prostate cancer is very low, indicating the possibility
of hypofractionation. Warning statements were published not to go below a few fractions in
order to maintain the benefit of reoxygenation. Yoshioka and colleagues were the first to test
high dose rate (HDR) brachytherapy as monotherapy for prostate cancer. The good results were
confirmed by others and recent years saw several successful attempts to treat prostate cancer
by a single HDR fraction.
The implications of this development for the understanding of the biological basis of
radiotherapy will be discussed.
23
Submitted abstracts
Three New Mechanistic Radiobiological Models for Cell Response to Radiation
Dževad Belkić & Karen Belkić: Department of Oncology-Pathology, Karolinska Institutet
Second-order kinetics are utilized to describe the three main mechanisms for surviving
fractions of cells after irradiation. These are a direct yield of lethal lesions by single event
inactivation, metabolic repair of radiation lesions and transformation of sublethal to lethal
lesions by further irradiations. The mass action law is employed to set up a system of time-
dependent, non-linear differential equations for average molar concentrations of the invoked
species, such as DNA substrates as lesions, enzyme repair molecules, a pool of intracellular
repair substances, etc. The analytical solutions of these coupled rate equations are derived.
Three new radiobiological models, based on the concept of a pool of repair molecules and
Michaelis-Menten enzyme catalysis, each having three dose-range biologically interpretable
parameters of clinical relevance, are analyzed from two different systems of rate equations
that are solved by extracting the concentration of lethal lesions whose time development is
governed by the said three mechanisms. The Poisson distribution of the derived
concentrations of lethal lesions gives the closed expression for the cell surviving fractions
after irradiation. Exploiting the derived asymptotes of the analytical solutions, the three new
dose-effect curves are shown to exhibit shoulders at intermediate doses preceded by the
exponential cell kill with a non-zero initial slope and followed by the exponential fall-off
with the reciprocal of the mean lethal dose as the final slope. All three dose regions are
universally as well as smoothly interconnected in the ensuing surviving fractions. The
proposed three radiobiological models are equally valid for low- and high-dose per fraction
treatment regimens. It is expected that these new models will find the most needed
applications in hypofractionation treatment schedules (stereotactic radiosurgery, stereotactic
body radiotherapy) for which the low-dose linear-quadratic model is known to be seriously
challenged.
24
Improved Distinction by MR Spectroscopy of Suspicious Lesions after
Radiation Therapy among Children with Primary Brain Tumors
Karen Belkić and Dževad Belkić
Department of Oncology-Pathology, Karolinska Institutet
Brain tumors are the leading cause of cancer-related deaths among children. Radiation
therapy (RT) is indicated for unresectable and/or malignant brain tumors. After RT, primary
brain tumors frequently recur. Changes in brain tissue can also be provoked by RT, and these
are difficult to distinguish from recurrent tumor using magnetic resonance imaging (MRI).
Increasing attention in pediatric neuro-oncology has been given to magnetic resonance
spectroscopy (MRS) and spectroscopic imaging (MRSI). We systematically examine MRS
and MRSI data concerning distinction of recurrent brain tumor versus radiation necrosis post-
RT among children with primary brain tumors. From the limited available data, it appears
that, among children, choline-to-creatine ratios > 2 indicate tumor recurrence whereas
choline-to-creatine ratios < 2 are associated with radiation necrosis. More extensive analysis
of the adult plus pediatric population, however, has indicated that, while helpful, metabolite
ratios do not provide unequivocal distinction between recurrent tumor and radiation necrosis.
These limitations are likely related to reliance upon sub-optimal signal processing methods
conventionally used within MRS and MRSI. We demonstrate that conventional Fourier
processing using clinical MR scanners may generate inaccurate metabolite ratios. Optimized
signal processing via the fast Padé transform not only yields accurate metabolite ratios with
very high resolution, but also provides reliable quantification of over 20 brain metabolites.
Padé-optimized MRS and MRSI could thereby facilitate differential diagnosis of new lesions
appearing on MRI among children treated for primary brain tumors. These advantages of the
fast Padé transform are more broadly applicable to MRS and MRSI within pediatric neuro-
oncology and beyond.
25
Gender-related differences in pathological and clinical tumour response based
on immunohistochemical proteins expression in rectal cancer patients treated
with short course of preoperative radiotherapy
Anna Gasinska1, Agnieszka Adamczyk1, Joanna Niemiec1, Beata Biesaga1, Zbigniew
Darasz2, Jan Skolyszewski3 1Department of Applied Radiobiology, 2Department of Surgery, 3Department of Radiation Oncology,
Oncology Center, Maria Sklodowska - Curie Memorial Institute, Cracow, Poland
Introduction Assessment of prognostic value of pretreatment expression of six proteins in
rectal cancer for early pathological tumour response (pTR), clinical tumour response (CTR)
to preoperative radiotherapy (RT), and the potential difference between these parameters
depending on patient gender.
Material and Methods 111 patients were treated with short preoperative course of RT
(SCRT) with 5 Gy dose per fraction during 5 days, followed by surgery 3 to 53 days (mean
21 days) later. Expression of GLUT-1, Ku70, BCL-2, P53 proteins was assessed
immunohistochemically. Tumour regression after RT was assessed by surgeons at the time of
operation (CTR), and by pathologist on the excised tumour mass (pTR).
Results There were 76 men and 35 women. There were 27 cTNM stage I, 69 stage II and 15
stage III tumours. We found 26 well-differentiated, 80 moderately-differentiated and 3
poorly-differentiated tumours. Significant differences in Ki-67, GLUT-1, Ku 70 and BCL-2
expression between male and female tumours were observed for pathological stage (pTNM)
and grade. Association between proteins expression and pTNM, pTR and CTR was analysed
separately for short (≤ 15 days) and long (> 15 days) break between RT and surgery and
males and female patients. For SCRT with short break none protein was significantly related
to pTNM, and for pTR higher Ki-67 and lower BCL-2 expression were correlated with pTR.
In the male subgroup, BCL-2 overexpression was predictive. For SCRT with long break none
of the proteins was predictive for pTR but Ki-67, Ku70 (in female subgroup) and BCL-2
expression were positively correlated with pTNM, what was shown for the first time. BCL-2
overexpression was associated with CTR in female subgroup, only.
26
Conclusion In SCRT long break in the treatment should be avoided because correlation
between Ki-67, KU70 and BCL-2 expression and pTNM after RT might indicate tumour
progression reflecting tumour cell repopulation.
27
An association between low-dose hyper-radiosensitivity
and the early G2-phase checkpoint
D. Słonina1, A. Gasińska1, A. Janecka1, B. Biesaga1, D. Kabat2 1Dept. of Applied Radiobiology, Centre of Oncology, Krakow, Poland
2Dept. of Medical Physics, Centre of Oncology, Krakow, Poland
The phenomenon of low-dose hyper-radiosensitivity (HRS) is an effect in which cells die
from excessive sensitivity to low doses (<0.3 Gy) of ionizing radiation but become more
resistant (induced radioresistance, IRR) to larger doses. The mechanisms regulating HRS/IRR
biology seem to be complicated and are continuously investigated. According to the proposed
hypothesis, HRS reflects the apoptotic death of cells as a consequence
of ineffective early G2-phase arrest of cells irradiated in G2 phase. According to some data
the checkpoint is activated after exposure to a threshold dose of 0.2 Gy and above.
In our previous study, using flow cytometry-based clonogenic survival assay, the HRS
response was demonstrated for normal fibroblasts (asynchronous and G2-phase enriched cell
populations) of 4 of the 25 cancer patients investigated (IJRBP 2014, 88, 369-376). This gave
us a reason to examine the mechanism underlying HRS phenomenon. The aim of our study
was to determine an association between low-dose hyper-radiosensitivity and the early G2-
phase checkpoint activation. The response of the checkpoint was examined by assessment of
the progression of irradiated cells into mitosis using the mitotic marker, phosphorylated
histone H3. The number of mitotic cells as a function of radiation dose was assessed for the
asynchronous fibroblasts of 4 HRS-positive and 4 HRS-negative patients.
After irradiation with doses lower than 0.3 Gy HRS-positive fibroblasts continued to enter
mitosis and, as a result, no reduction in mitotic ratio was observed. The dose over which the
number of mitotic cells started to decrease corresponded well with the transition dose
(specific for each HRS-positive patient) required to overcome HRS. In contrast,
HRS-negative fibroblasts exhibited an immediate post-irradiation decrease in the mitotic ratio
even at the lowest doses.
The data provides support to the existing hypothesis suggesting the relationship between the
HRS response and failure of the early G2-phase checkpoint.
29
Aberrant DNA damage response and receptor tyrosine kinase signaling in non-
small cell lung cancer tumor initiating cells
Lovisa Lundholm, Petra Hååg, Dali Zong, Therese Juntti, Birgitta Mörk, Rolf Lewensohn
and Kristina Viktorsson
Karolinska Biomics Center, Department of Oncology-Pathology, Karolinska Institutet and Karolinska
University Hospital, Stockholm, Sweden.
Increasing evidence suggest that tumor initiating cells (TICs), also called cancer stem cells
(CSC), are partly responsible for resistance to DNA-damaging treatment. Using a model of
sphere-forming non-small cell lung cancer (NSCLC) cells to characterize the specific
properties of TICs, we show that they displayed reduced apoptotic response and less
pronounced cell cycle arrest after DNA-damaging treatment e.g. irradiation (IR) and
cisplatin. TICs also showed altered DNA repair signaling with diminished DNA damage-
induced phosphorylation of ATM and KAP1. Receptor tyrosine kinases (RTK) are important
regulators of intracellular signal transduction pathways which influence tumor cell sensitivity
to DNA damaging treatments. Here we examined if NSCLC TICs show an altered basal and
IR-induced activation of 28 RTKs and 11 important signaling nodes as compared to bulk
tumor cells using PathScan RTK Signaling Antibody Array Kit. TICs displayed decreased
basal phosphorylation of insulin-like growth factor 1 receptor (IGF-1R) and signal transducer
and activator of transcription 1 (STAT1) (Tyr701) as compared to bulk cells. Total IGF-1R
levels were significantly lower in NSCLC TIC as compared to bulk cells. A higher degree of
extracellular signal-regulated kinase (ERK) phosphorylation was evident in TIC and
concordantly MEK inhibition reduced TIC viability. Moreover, MEK inhibition also
decreased clonogenicity upon IR suggesting that MEK and downstream signaling impart on
TIC radiation response. In conclusion, we demonstrate that NSCLC TICs have both altered
DNA damage signaling response as well as altered RTK activation which may influence their
resistance to DNA damaging treatments such as IR.
30
The influence of repopulation mechanisms on treatment gap timing in head and
neck cancer radiotherapy
Loredana G. Marcu1,2 1Faculty of Science, University of Oradea, 410087, Romania, 2School of Chemistry & Physics,
University of Adelaide, SA 5000, Australia
email: [email protected]
Background: In head and neck cancers tumour repopulation and hypoxia are the leading
cause of treatment failure. Various mechanisms behind repopulation have been identified:
recruitment of quiescent cells, accelerated stem cell division, loss of asymmetrical division of
stem cells and abortive division. The latter mechanism refers to the proliferation of
differentiated cells rather than stem cells, however, it can still contribute towards overall
repopulation. Advanced head and neck cancers are often managed with altered fractionation
schedules, such as accelerated radiotherapy, in order to overcome malignant repopulation.
Treatment breaks can be employed to allow normal tissue recovery. The aim of the present
work is to illustrate the influence of repopulation due to both stem and differentiated cells on
treatment break timing during accelerated radiotherapy in head and neck carcinomas.
Methods: A Monte Carlo computational method has been employed to simulate the growth
of a head and neck carcinoma, with biologically realistic parameters. In the model, stem and
differentiated cells represent about 16% of the tumour population and have a mean cell cycle
time of 33h. The aim of the simulation was to follow the behaviour of the virtual tumour on a
temporal scale and to analyse tumour response to altered fractionation radiotherapy when all
the mechanisms responsible for repopulation are activated. Based on a RTOG schedule (1.6
Gy/fraction, twice daily, 6 hours apart, 5 days a week and a total number of 42 fractions),
three different timings for treatment interruptions have been simulated using the Linear
Quadratic model: after 20, 24 and 28 fractions, respectively. Both stem and differentiated cell
have been monitored and their contribution towards tumour growth analysed and discussed.
Results: Due to the activation of repopulation mechanisms during radiotherapy, there is a
large increase in the percentage of stem and also differentiated cells that contribute to tumour
development. While before radiotherapy the tumour consisted of 6% stem cells and 10%
differentiated cells, during accelerated radiotherapy, both percentages increased drastically,
depending also on the timing of treatment breaks. Therefore, the average percent of stem cells
varied from 41.3% (break after 20 fractions) to 36.6% (break after 28 fractions), while the
31
average percent of differentiated cells varied from 30.5% (break after 20 fractions) to 33.7%
(break after 28 fractions). An interesting observation is the fact that the percentage of stem
cells decreases with the delay of treatment gap. This is because early breaks (after 20
fractions) do not allow sufficient cell kill among the continuously proliferating stem cells to
control the tumour. The behaviour of differentiated cells is just the opposite, to keep a
constant cell kill along the treatment.
Conclusions: The model has shown that the timing of treatment breaks is an important factor
influencing tumour control in rapidly proliferating tissues. Not only stem cells but also
differentiated cells, via abortive division, can contribute to malignant cell repopulation during
treatment. Differentiated cells undergoing abortive division are ‘doomed’ cells as they
eventually cease creating new cells and die out. On the other hand, stem cells are able to
regrow the tumour, thus for their eradication there is need for fine adjustments of treatment
parameters.
32
Cluster pattern analysis of energy deposition sites for protons, other light ions
and brachytherapy sources
Fernanda Villegas1, Gloria Bäckström1, Nina Tilly1,2, and Anders Ahnesjö1 1Department of Radiology, Oncology, and Radiation Science, Section of Medical Radiation
Physics, Uppsala University. Sjukhusfysik Ing 81. Akademiska Sjukhuset, SE-75185
Uppsala, Sweden, 2Elekta Instrument AB, Box 1704, 75147, Uppsala, Sweden
[email protected], [email protected],
[email protected], [email protected]
The spatial pattern of energy deposition (ED) sites formed by the interactions of ionizing
particles and their secondary electrons within a cell nucleus could determine the severity of
the initial damage to the DNA and hence influence the relative biological effect (RBE) of the
actual radiation type. RBE data is needed to improve the accuracy of treatment planning due
to the many concurrent radiation qualities usually characterized by their linear energy transfer
(LET) which is an imprecise predictor of RBE. It could be advantageous if RBE could be
predicted in a reliable way. The aim of this work was to analyse the cluster patterns of EDs
formed by proton and other light ions as well as five of the most frequently used
brachytherapy sources, 103Pd, 125I, 192Ir, 137Cs and 60Co. The ED data was obtained by
performing event-by-event Monte Carlo simulations with the LIonTrack and PENELOPE
radiation transport codes. A novel cluster method was developed which does not bias
radiation clustering properties towards preselected cluster scoring volumes. The number of
clusters with cluster order (number of EDs contained in the cluster) larger than two increases
with decreasing atomic number for ions with the same LET (25.7 eV nm-1). The ratio of
cluster order distributions for the brachytherapy sources reveals that the two lower mean
photon energy sources (103Pd, and 125I) differ by as much as 15% with respect to 60Co,
while the ratio differs by only 2% for the two sources with higher mean photon energy
(192Ir, and 137Cs). Our results encourage the use of cluster order distributions as a
foundation for RBE analysis since it can discern differences between ion species not
predicted by LET variations.
33
A study of V79 cell survival after for proton and carbon ion beams as
represented by the parameters of Katz’ track structure model.
L. Grzanka(1), M. P. R. Waligórski (1,2), and N. Bassler(3)
1 Institute of Nuclear Physics, Polish Academy of Sciences, Kraków, Poland, 2 The Marie-
Skłodowska-Curie Centre of Oncology, Kraków Division, Kraków, Poland, 3. Department of Physics
and Astronomy, Aarhus University, Aarhus, Denmark
Katz’s theory of cellular track structure (1) is an amorphous analytical model which applies a
set of four cellular parameters representing survival of a given cell line after ion irradiation.
Usually the values of these parameters are best fitted to a full set of experimentally measured
survival curves available for a variety of ions. Once fitted, using these parameter values and
the analytical formulae of the model calculations, cellular survival curves and RBE may be
predicted for that cell line after irradiation by any ion, including mixed ion fields. While it is
known that the Katz model parameters fitted to heavier ion data may yield unsatisfactory
predictions of proton response, to our knowledge, no comprehensive data set which includes
proton and heavier ion irradiations, measured in one laboratory, has been published. To study
the consistency of evaluating parameters of this model from different sets of data obtained for
the same cell line and different ions, measured at different laboratories, we have fitted model
parameters to a set of carbon-irradiated V79 cells, published by Furusawa et al. (2), and to a
set of proton-irradiated V79 cells, published by Wouters et al. (3), separately. We found that
values of model parameters best fitted to the carbon data of Furusawa et al. yielded
predictions of V79 survival after proton irradiation which did not match the V79 proton data
of Wouters et al. Fitting parameters to both sets combined did not improve the accuracy of
model predictions of the proton response. This suggests that for increased accuracy of a
therapy planning system based on Katz’s model, different sets of parameters may need to be
used to represent cell survival after proton irradiation from those representing survival of this
cell line after heavier ions, up to and including carbon irradiation.
1. Katz, R., Track structure in radiobiology and in radiation detection. Nuclear Track
Detection 2: 1-28 (1978).
2. Furusawa Y. et al. Inactivation of aerobic and hypoxic cells from three different cell lines
by accelerated 3He-, 12C- and 20Ne beams. Radiat Res. 2012 Jan; 177(1):129-31.
34
3. Wouters BG et al. Measurements of relative biological effectiveness of the 70 MeV proton
beam at TRIUMF using Chinese hamster V79 cells and the high precision cell sorter assay.
Radiat Res.1996 Aug. 146(2): 159-170.
35
Genetic variants and expression of AEG-1 in relation to clinical outcome and
radiotherapy response in colorectal cancer patients and cell lines
Sebastian Gnosa1, Veronica Patcha Brodin1, Ivana Ticha1, Gunnar Adell1, Gunnar Arbman3, Hong
Zhang2, Xiao-Feng Sun1
1Division of Oncology, Department of Clinical and Experimental Medicine, Faculty of Health
Sciences, County Council of Östergötland, University of Linköping, Sweden, 2School of Medicine,
Örebro University, SE-70128 Örebro, Sweden, 3Department of Surgery, Vrinnevi Hospital,
Norrköping, Sweden
Background: Colorectal cancer (CRC) is the third most common cancer in men and the
second in woman worldwide. The therapy of CRC has been critically improved during the
past two decades, but the treatment response varies significantly between different treatments
and patients. Therefore it is of need to search for biomarkers for more suitable prognosis and
treatment. The aim of this study was to investigate genetic variants of the astrocyte elevated
gene-1 (AEG-1), its expression in CRC patient samples and colon cancer cell lines and the
potential correlation to clinicopathological variables and treatment response.
Materials and Methods: We analysed genetic variants and the mRNA and protein
expression of AEG-1 in 593 CRC patient samples by direct Sanger sequencing, qPCR and
immunohistochemistry. Of the patients, 158 rectal cancer patients were enrolled in the
Swedish clinical trial of preoperative radiotherapy (RT). AEG-1 expression in response to γ-
irradiation was analysed by Western blot in 5 colon cancer cell lines. We inhibited the AEG-1
expression by siRNA in the cell lines, and their survival was analysed after γ-irradiation.
siRNA in the cell lines, and their survival was analysed after γ-irradiation.
Results: By direct sequencing we found 29 variants, of which 12 were novel. Six of the
variants were exonic and 23 intronic. The variant c.1353G>A, rs2331652 was found only in 4
samples and the variant c.1679-6 T>C in only one sample. In the cell lines KM12C,
KM12SM and KM12L4a we found a deletion in exon 4 (c.595_598delAGAG). AEG-1
mRNA and protein expression revealed a significant increased AEG-1 expression in the
primary tumors and metastases compared to the normal mucosa in all patient cohorts
(p<0.006) and on the mRNA level a higher expression in tumors located in the rectum
compared to tumors in the colon (p=0.047). In the rectal cancer patients from the Swedish
trail of preoperative RT, a high AEG-1 expression correlated with a higher risk of distant
recurrence and disease free survival (p=0.001 respectively) independently of the tumor stage,
only in patients receiving preoperative RT. AEG-1 knock-down in the radioresistant cell lines
36
KM12L4a, SW480 and SW620 resulted in reduction of the survival after radiation, but not in
the radiosensitive cell lines KM12C and HCT116. The AEG-1 expression was up-regulated
shortly after 2 Gy γ-radiation in KM12C, KM12L4a and SW480, but not in SW620 and
HCT116.
Conclusion: We found several novel AEG-1 variants. The AEG-1 expression at the mRNA
and/or protein level was up-regulated in the primary tumour and metastases compared to the
normal mucosa. Furthermore we showed for the first time that the AEG-1 expression is an
independent prognostic factor for distant recurrence and disease-free survival in rectal cancer
patients after treatment with preoperative RT. The cell line studies suggest AEG-1 as a
promising radiosensitizing target for rectal cancer.